| Title | MDL Patch Correspondences on Unlabeled Images with Occlusions |
| Authors | Johan Karlsson, Karl Åström |
| Alternative Location | http://dx.doi.org/10.1109/C... |
| Publication | 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPR Workshops) |
| Year | 2008 |
| Pages | 999 - 1006 |
| Document type | Conference paper |
| Conference name | 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPR Workshops) |
| Conference Date | 2008-06-23 - 2008-06-28 |
| Conference Location | Anchorage, Alaska, USA |
| Status | Published |
| Quality controlled | Yes |
| Language | eng |
| Abstract English | Automatic construction of Shape and Appearance Models from examples <br> via establishing correspondences across the training set has been successful in the last decades.<br> One successful measure for establishing correspondences of high quality is minimum description length (MDL).<br> In other approaches it has been shown that parts+geometry models which model the appearance of parts of the object and the geometric relation between the parts<br> have been successful for automatic model building.<br> In this paper it is shown how to fuse the above approaches and use MDL to<br> fully automatically build optimal parts+geometry models from unlabeled images. |
| Keywords | computational geometry, image processing, MDL patch correspondence, unlabeled images, occlusions, automatic construction, shape model, appearance model, training set, minimum description length, automatic model building, |
| ISBN/ISSN/Other | ISBN: 978-1-4244-2339-2 |
Questions: webmaster
Last update: 2013-04-11
Centre for Mathematical Sciences, Box 118, SE-22100, Lund. Telefon: +46 46-222 00 00 (vx)